PCBUILD Archives

Personal Computer Hardware discussion List

PCBUILD@LISTSERV.ICORS.ORG

Options: Use Forum View

Use Monospaced Font
Show Text Part by Default
Show All Mail Headers

Message: [<< First] [< Prev] [Next >] [Last >>]
Topic: [<< First] [< Prev] [Next >] [Last >>]
Author: [<< First] [< Prev] [Next >] [Last >>]

Print Reply
Subject:
From:
David Gillett <[log in to unmask]>
Reply To:
PCBUILD - Personal Computer Hardware discussion List <[log in to unmask]>
Date:
Wed, 7 Dec 2005 23:28:40 -0800
Content-Type:
text/plain
Parts/Attachments:
text/plain (49 lines)
On 6 Dec 2005 at 0:53, Venkat Viswanathan wrote:

> HI all members,
>
> I have a celeron D 2.4 GHz. recently i did a md5sum check on a knoppix
> live cd download (700 MB approx) and i got the "blinking cursor" for
> what i thought was a long time so i thought something was wrong and
> pressed "enter" twice. after about 30 seconds the results of the check
> came through fine.
>
> out of curiosity i did the check again and timed it. it came to
> 2mts.50 secs.
>
> my question is (1)if the processor was p4 2.4 GHZ instead of celeron
> would the results have come out faster? on a side note would a 3.0 GHz
> processor been faster.
>
> TIA and good day---Venkat

  Maybe not.

  To answer either question, we have to see how the differences between the
options relate to the computing task to be performed.

  MD5 is definitely a processor-intense calculation, to be sure, which
should give a higher-clock-rate CPU an edge.  But over a 700MB image, it's
entirely likely that the disk read bandwidth is the bottleneck rather than
the CPU, and so a faster CPU might only be faster at waiting for the hard
drive.
  The P4's hyperthreading feature is going to be most useful in multi-
tasking environments; it's more likely to reduce the extent to which the MD5
calculation slows down every other task on the machine, that to actually
speed up the single MD5 task.
  I think there's also more cache on the P4 than the Celeron, but cache only
helps if you're referring to the same data repeatedly.  In this case, the
calculation needs to read every part of the 700 MB *once* -- better caching
will never get a chance to make any difference.

  There is *one* way in which better/more cache may help, and that is if the
entire inner calculation code for the MD5 algorithm fits into the cache, so
that only the image data needs to be read.  But without more detail, I don't
know that isn't already happening on the Celeron.

David Gillett

              The NOSPIN Group is now offering Free PC Tech
                     support at our newest website:
                          http://freepctech.com

ATOM RSS1 RSS2